Visualization of ultrasonic-guided-wave propagation behaviors in human long bone

Guided wave ultrasound technology is widely recognized for its applications in non-destructive testing. The technology is being explored for its potential use in bone tissue characterization and imaging to evaluate bone health and fracture risk. Cortical bone with complex microstructure introduces significant dispersion and attenuation effects on ultrasonic guided waves (UGWs). Therefore, it is crucial to comprehend the frequency-dependent propagation behaviors of co-excited wave modes in order to study UGW propagation and develop wave-based approaches for bone assessment. This research employs state-of-the-art signal processing methods, including estimation of signal parameters by iterative rational approximation (ESPIRA) and superlet time–frequency analysis, to quantitatively visualize the dispersive characteristics of UGWs measured from a human tibia using axial transmission ultrasound. An ESPIRA-based matrix pencil algorithm simultaneously extracts dispersive modal wavenumber and attenuation from experimental UGW signals. The estimation of dispersion relations is formulated as a generalized eigenvalue problem of Loewner matrices, which enhances numerical stability and noise resistance. To validate the estimation accuracy and identify the propagating wave modes, the computed dispersive features are compared with numerically-modeled dispersion curves by semi-analytical finite-element simulation. Using bone UGW signals as a case study, the proposed methodology can be applied to other UGW analyses beyond bone quantitative ultrasound.